🔭 astrophysics

Cross-correlation of SPT-3G D1 CMB lensing and DES Y3 galaxy lensing

This paper presents the first high-significance ( $\sim 14\sigma$ ) measurement of the cross-correlation between SPT-3G D1 CMB lensing and DES Y3 galaxy lensing, yielding a robust constraint on the matter clustering amplitude $S_8$ that is consistent with Planck and DES shear-only results while providing competitive limits on intrinsic alignments and baryonic feedback.

Original authors: A. Ouellette, Y. Omori, E. Anderes, A. J. Anderson, B. Ansarinejad, M. Archipley, L. Balkenhol, D. R. Barron, K. Benabed, A. N. Bender, B. A. Benson, F. Bianchini, L. E. Bleem, S. Bocquet, F. R. Bouch

Published 2026-06-26

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: A. Ouellette, Y. Omori, E. Anderes, A. J. Anderson, B. Ansarinejad, M. Archipley, L. Balkenhol, D. R. Barron, K. Benabed, A. N. Bender, B. A. Benson, F. Bianchini, L. E. Bleem, S. Bocquet, F. R. Bouchet, E. Camphuis, M. G. Campitiello, J. E. Carlstrom, J. Carron, C. Chang, C. L. Chang, P. M. Chichura, A. Chokshi, T. -L. Chou, A. Coerver, T. M. Crawford, C. Daley, T. de Haan, K. R. Dibert, M. A. Dobbs, M. Doohan, D. Dutcher, C. Feng, K. R. Ferguson, N. C. Ferree, K. Fichman, A. Foster, S. Galli, A. E. Gambrel, A. K. Gao, F. Ge, F. Guidi, S. Guns, N. W. Halverson, E. Hivon, G. P. Holder, W. L. Holzapfel, J. C. Hood, A. Hryciuk, N. Huang, T. Jhaveri, F. Kéruzoré, A. R. Khalife, L. Knox, K. Kornoelje, C. -L. Kuo, K. Levy, Y. Li, A. E. Lowitz, C. Lu, G. P. Lynch, T. J. Maccarone, A. S. Maniyar, E. S. Martsen, F. Menanteau, M. Millea, J. Montgomery, Y. Nakato, T. Natoli, Z. Pan, P. Paschos, K. A. Phadke, A. W. Pollak, K. Prabhu, W. Quan, M. Rahimi, A. Rahlin, C. L. Reichardt, M. Rouble, J. E. Ruhl, A. C. Silva Oliveira, A. Simpson, J. A. Sobrin, A. A. Stark, J. Stephen, C. Tandoi, C. Trendafilova, J. D. Vieira, A. G. Vieregg, A. Vitrier, Y. Wan, N. Whitehorn, W. L. K. Wu, M. R. Young, J. A. Zebrowski

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the universe as a giant, invisible ocean of matter. We can't see this matter directly, but we can see how it bends the light from things behind it, much like how a funhouse mirror distorts the reflection of a person standing in front of it. This bending of light is called gravitational lensing.

Scientists have been trying to map this invisible ocean for decades using two different "cameras":

The Galaxy Camera: This looks at billions of distant galaxies. As their light travels to us, the invisible matter bends it, making the galaxies look slightly stretched or squished. This is called "cosmic shear."
The CMB Camera: This looks at the Cosmic Microwave Background (CMB), which is the oldest light in the universe, like a baby picture of the cosmos. As this ancient light travels to us, the invisible matter also bends it. This is called "CMB lensing."

The Problem:
Each camera has its own blind spots and quirks. The Galaxy camera is great at seeing nearby structures but gets confused by the shapes of the galaxies themselves. The CMB camera sees the whole history of the universe but gets "foggy" because of dust and other bright objects in space that look like the lensing signal.

The Solution: Cross-Checking
This paper is about taking the data from both cameras and comparing them, like two detectives comparing their notes on the same crime scene. If both cameras see the same distortion in the same place, we can be sure it's real and not a glitch in the equipment.

What They Did:
The researchers combined data from two massive projects:

SPT-3G: A telescope at the South Pole that looks at the CMB.
DES (Dark Energy Survey): A camera on a telescope in Chile that looks at galaxies.

They looked at a patch of sky about 1,300 square degrees (roughly the size of 5,000 full moons) where the views of both telescopes overlap.

The Big Breakthrough:
Usually, when looking at the CMB, the "fog" of dust and other space objects makes it hard to get a clean picture. To solve this, the team used a special trick: they only used the polarization (the direction the light waves are vibrating) of the CMB data.

Think of it like wearing polarized sunglasses. Regular sunglasses block some glare, but polarized ones are specifically designed to block the specific type of glare coming from water or roads. Similarly, by only looking at the polarized light, the scientists effectively "blocked out" the cosmic dust and other foreground noise that usually messes up the CMB lensing maps.

The Results:

High Confidence: They detected the connection between the two lenses with extremely high confidence (about 14 times stronger than random noise). This is the first time this has been done with such high clarity using only the "polarized" view.
Checking the Theory: They measured how clumpy the universe is (a number called $S_8$ ). Their result matched perfectly with what we expect from the Big Bang theory (Planck data) and what other galaxy surveys have found. This is a huge "thumbs up" for our current understanding of the universe.
Cleaning the Data: They tested different ways to clean the data. They found that while using all the data (including temperature) gives a slightly sharper picture, it also introduces a little bit of "noise" from the cosmic dust. However, their "polarization-only" method proved to be the cleanest and most reliable way to avoid those errors.

Why It Matters:
This study is like a rigorous quality control check. By showing that two completely different ways of looking at the universe (galaxies vs. ancient light) agree with each other, the scientists confirm that our models of how the universe grows and evolves are on the right track. It also proves that using "polarized sunglasses" on the CMB is a powerful new tool for future astronomers to get even clearer pictures of the invisible universe.

Technical Summary: Cross-correlation of SPT-3G D1 CMB lensing and DES Y3 galaxy lensing

Problem and Motivation
Weak gravitational lensing of galaxies (cosmic shear) and the Cosmic Microwave Background (CMB) provides direct probes of the cosmic matter density field. However, these observables are sensitive to different spatial scales, redshift ranges, and survey systematics. A persistent mild tension exists in the inferred value of the matter clustering amplitude parameter, $S_8 \equiv \sigma_8 \sqrt{\Omega_m/0.3}$ , where cosmic shear surveys typically measure values $\sim 2$ to $3\sigma$ lower than predictions based on primary CMB measurements (e.g., Planck). While CMB lensing alone is consistent with primary CMB, the combination of these independent datasets offers a unique opportunity to address this tension. Their cross-correlation is expected to be sensitive to scales straddling the nonlinear transition, allowing for simpler modeling than cosmic shear alone while retaining sensitivity to small-scale physics. Furthermore, cross-correlating independent datasets with different systematics provides a robust consistency test. The primary challenges in this measurement are contamination from galaxy intrinsic alignments (IA) and extragalactic foregrounds in CMB data (such as the Sunyaev-Zeldovich effect and Cosmic Infrared Background) that can mimic lensing signals.

Methodology
The authors present a measurement of the cross-correlation between CMB lensing convergence ( $\kappa_{\text{CMB}}$ ) and cosmic shear ( $\gamma$ ) over an overlapping sky area of approximately $1,300 \text{ deg}^2$ .

Data: The analysis utilizes new CMB lensing maps derived from the South Pole Telescope (SPT-3G) D1 dataset (temperature and polarization observations from 2019–2020) and the Dark Energy Survey Year 3 (DES Y3) shear catalogs.
Lensing Estimators: To mitigate foreground biases, the authors employ five different quadratic estimator (QE) variants:
1. GMV (Global Minimum Variance): Uses both temperature and polarization; lowest noise but most susceptible to foregrounds.
2. Pol (Polarization-only): Uses polarization data only; expected to be robust against extragalactic foregrounds (which are unpolarized) but has higher noise on small scales.
3. GMVprof (Profile-hardened): Mitigates bias from specific source profiles (e.g., clusters, radio sources).
4. GMVtszdpj (tSZ-deprojected): Removes thermal Sunyaev-Zeldovich (tSZ) contamination via gradient cleaning.
5. GMVxilc (Cross-ILC): Removes both tSZ and Cosmic Infrared Background (CIB) contamination.
Shear Samples: Two galaxy samples are used: the full DES Y3 sample ("Full") and a high-purity subsample of blue, star-forming galaxies ("Blue"), which is expected to have negligible intrinsic alignment amplitudes.
Analysis Pipeline: Cross-spectra ( $C_{\ell}^{\kappa\gamma}$ ) are measured using NaMASTER with pseudo- $C_\ell$ techniques. The analysis marginalizes over uncertainties in shear calibration, photometric redshifts, intrinsic alignments (using the Non-Linear Alignment model), and baryonic feedback (via the HMCode $\log T_{\text{AGN}}$ parameter).
Validation: The pipeline is validated using the Agora simulations, which model correlations between CMB lensing, galaxy lensing, and extragalactic foregrounds.

Key Contributions and Results

High-Significance Detection: The paper reports the first high-significance ( $\sim 14\sigma$ ) measurement of the CMB lensing–cosmic shear cross-correlation using a polarization-only CMB lensing reconstruction. This effectively sidesteps biases from extragalactic foregrounds. When including temperature data with foreground mitigation, the significance rises to $\sim 15\text{--}16\sigma$ .
Cosmological Constraints: Assuming a flat $\Lambda$ CDM cosmology and marginalizing over nuisance parameters, the fiducial analysis (Full $\times$ Pol) yields a constraint on the matter clustering amplitude of:
$S_8 = 0.833^{+0.047}_{-0.061}$
This result is consistent with both primary CMB results (Planck) and shear-only results (DES Y3).
Foreground Bias Characterization: The study quantifies foreground-induced biases. The raw GMV estimator (without foreground mitigation) shows a $\sim 20\text{--}40\%$ negative bias on large scales and a positive bias on small scales, consistent with tSZ contamination. The polarization-only and foreground-mitigated estimators show no significant bias relative to the true signal in simulations.
Intrinsic Alignments: Using the blue galaxy subsample with an informative prior on IA amplitude (motivated by external constraints), the constraint on $S_8$ tightens to $0.860 \pm 0.040$ . The analysis finds no evidence for non-zero IA in the full sample ( $A_1 = 0.21^{+0.98}_{-0.75}$ ) and no evidence for redshift evolution of the IA amplitude.
Joint Analysis with Planck: Combining the cross-correlation with Planck 2018 likelihoods allows for tighter constraints on astrophysical parameters. The joint analysis yields constraints on the IA amplitude ( $A_1$ ) competitive with shear-only analyses and provides a lower limit on baryonic feedback strength ( $\log T_{\text{AGN}} > 7.70$ ) when using the high-SNR profile-hardened estimator.

Significance
The paper claims significance primarily as a demonstration of the power of low-noise CMB polarization observations for cosmological cross-correlations. By achieving a $14\sigma$ detection using polarization-only data, the authors demonstrate a method to robustly avoid extragalactic foreground biases that plague temperature-based reconstructions. This work represents a significant improvement over previous polarization-only cross-correlation measurements (e.g., POLARBEAR $\times$ HSC at $3.5\sigma$ ). The results provide a consistency check between CMB and galaxy lensing probes, confirming that the $S_8$ tension is not driven by a fundamental mismatch between these specific datasets. Furthermore, the study validates the use of tailored data selections (blue galaxies) and foreground mitigation techniques to isolate cosmological signals, setting a foundation for future analyses with larger datasets from LSST, Euclid, and the Simons Observatory.

More like this